How to Improve a PSLE Science Investigation Without Changing the Scientific Question

Wait, What? A “Better” Experiment Can Accidentally Become a Different Experiment

A learner is asked to improve a Science investigation. The original question is simple: How does the amount of light affect the growth of a plant?

The learner proposes an improvement: “Use different plant species so we can see more clearly.”

That may create more interesting data. But it has also changed the question. The investigation is no longer testing only the effect of light. It is now mixing light level with plant type.

A method improvement should strengthen the evidence for the original scientific question. It should not quietly replace that question with a new one.

This distinction matters because many PSLE Science method questions tempt learners to add more apparatus, more conditions or more steps simply because those additions sound “more scientific”. The strongest answer is not the one with the most equipment. It is the one that repairs the actual weakness while preserving the scientific relationship being tested.

Quick Answer

Before changing any investigation method, freeze the original question in one sentence:

How does ______ affect ______ under ______ conditions?

Then check every proposed improvement against that sentence. A good repair may improve control, measurement, repetition, sampling, timing or observation quality. But the changed factor, the outcome being measured and the core relationship should remain aligned with the original question unless the question itself is intentionally being redesigned.

Use this route:

STATE THE ORIGINAL QUESTION → IDENTIFY THE CHANGED FACTOR → IDENTIFY THE MEASURED OUTCOME → NAME THE METHOD WEAKNESS → PROPOSE THE SMALLEST REPAIR → CHECK THAT THE REPAIR DOES NOT ADD A NEW CAUSE OR REPLACE THE OUTCOME → KEEP OTHER RELEVANT CONDITIONS COMPARABLE → RUN THE EVIDENCE CHECK → STATE WHAT THE IMPROVED METHOD CAN NOW SUPPORT.

The Exact PSLE Science Learning Job This Guide Owns

This guide owns one learner job: how a Primary 5 or Primary 6 learner improves a PSLE Science investigation while preserving the original scientific question, changed factor and intended measured outcome.

It does not replace the general guides on fair tests, variables, repetition, sampling or measurement. Those owners explain their own mechanisms. This page owns the decision that sits above them:

Does my improvement make the evidence stronger for the same question, or have I accidentally changed what is being tested?

Why This Matters in the Current 2026 PSLE Science Frame

For examination from 2026, PSLE Science assesses attainment in the 2023 Primary Science syllabus. The official assessment objectives include knowledge with understanding and application of knowledge and scientific inquiry. Scientific inquiry includes making predictions and formulating hypotheses, interpreting and analysing information, evaluating observations, information and methods, and communicating explanations and reasoning.

Method evaluation therefore is not a checklist exercise. The learner is expected to judge whether the procedure can actually produce evidence for the scientific relationship under investigation.

The Question Signature: Three Things You Must Protect

Every simple investigation has a question signature. At Primary level, three parts are especially important:

Changed factor: what the investigator deliberately changes.
Measured outcome: what is observed or measured to see the effect.
Comparison conditions: the relevant conditions that must stay comparable so another cause does not take over.

Example:

Question: How does exposed surface area affect the amount of water lost from a wet material over the same time?

Changed factor: exposed wet surface area.
Measured outcome: amount of water lost.
Important comparable conditions: starting water amount, time, surroundings and material where relevant.

A method improvement should protect that architecture.

What Counts as a Genuine Improvement?

A genuine improvement reduces an evidence weakness without changing the target relationship.

Weakness	Possible repair	Original question preserved?
One trial may be unusual	Repeat the same planned trial	Yes
One specimen may be unrepresentative	Use several suitable similar specimens	Yes, if specimen selection does not introduce a new factor
Measurement is too coarse	Use a more suitable instrument or scale	Yes
Observation categories are vague	Define a consistent observation criterion	Yes
Another variable also changes	Keep that variable comparable	Yes
Only one test value is used	Test more values of the same changed factor	Usually yes
The learner changes the measured outcome to something easier	Different evidence target	No, unless it is a valid operational measure of the same intended outcome
The learner changes plant species, material type and light level together	Adds new changed factors	No

The Smallest-Repair Rule

When several improvements are possible, prefer the smallest change that fixes the identified weakness.

Why?

It keeps the original question visible.
It reduces the chance of creating new uncontrolled variables.
It makes cause-and-effect reasoning easier.
It lets you explain exactly why the repair strengthens the evidence.

“Use a completely different apparatus” may be unnecessary if the real problem is only that the ruler markings are too coarse.

Worked Example 1 — Fix the Measurement, Not the Question

Question: How does the length of a stretched elastic band affect the distance travelled by a toy car?

The learner notices that distance is measured with a ruler marked every 5 cm, making small differences hard to detect.

Good repair: use a ruler or measuring tape with finer suitable divisions while keeping the same car, release method, surface and stretched-band conditions.

Bad repair: change to a heavier toy car “so the distances are easier to compare”.

Why is the second repair poor? Car mass may also affect the outcome. The improved experiment would no longer isolate the original relationship as cleanly.

Worked Example 2 — More Test Values, Same Question

Question: How does water temperature affect the time taken for a stated change to occur?

The original test uses only 20°C and 60°C.

If the learner wants to understand the relationship across the range, testing 30°C, 40°C and 50°C can strengthen the evidence while preserving the same changed factor and measured outcome.

However, changing the substance at each temperature would create a new factor. More data are useful only if they belong to the same scientific question.

Worked Example 3 — Repetition Without Question Drift

Question: Which of two surfaces allows the same toy car to travel farther from the same release condition?

One run on each surface produces a difference.

Good repair: repeat the same procedure several times for each surface using the same intended release rule.

Bad repair: use several different cars because “more objects make it more reliable”.

Different cars can differ in mass, wheel condition and friction. Unless the scientific question is about cars generally and the design handles that variation deliberately, changing cars may blur the original surface comparison.

Worked Example 4 — Sampling Can Preserve or Change the Question

Question: How does a stated environmental condition affect the growth of seedlings of the same type?

Using several suitable similar seedlings in each condition may strengthen the evidence by reducing dependence on one unusual plant.

But using four different plant species in one condition and four other species in another changes the scientific object. Now species differences can explain the outcome.

Worked Example 5 — A Control Improvement That Keeps the Target

Question: Does material type affect how quickly equal amounts of hot water cool?

Two containers use different wrapping materials, but one also begins with more water.

Repair: use equal starting amounts of water and the same starting temperature, while changing only the wrapping material.

The repair does not change the question. It removes a competing explanation.

Worked Example 6 — Changing the Outcome Can Change the Question

Question: How does light level affect plant growth over one week?

The original outcome is change in height. A learner proposes measuring leaf colour instead because it is easier.

Leaf colour may be scientifically interesting, but it is not the same outcome as growth unless the question explicitly defines growth through a different valid measure.

This is a common investigation mistake: choosing an easy measurement that does not answer the original question.

Operational Measures: When a New Measurement Still Preserves the Question

Sometimes changing the measurement method does not change the outcome.

Suppose the question concerns the distance a car travels. Replacing a coarse floor marker with a measuring tape changes the method of measurement, not the scientific outcome.

Likewise, replacing a vague brightness description with a suitable light sensor may preserve the question if brightness is still the intended outcome and the sensor is appropriate.

The test is:

Does the new method measure the same scientific outcome more clearly, or has it replaced that outcome with a different property?

Five Types of Method Repair

1. Control Repair

Keep a competing factor comparable so the changed factor remains interpretable.

2. Measurement Repair

Use a more suitable method, scale or criterion to observe the intended outcome.

3. Repetition Repair

Repeat the same planned test to see whether the result is reasonably consistent.

4. Sampling Repair

Use more suitable similar specimens when natural variation makes one specimen too fragile a basis for a broader claim.

5. Range or Timing Repair

Test more values of the same changed factor or measure at time points that can reveal the feature the question asks about.

Each repair should leave the question signature intact.

How a Repair Accidentally Changes the Changed Variable

Suppose the original question changes only surface roughness.

If the learner “improves” the test by choosing different surfaces made from different materials and with different slopes, the investigation now changes several things at once.

A method improvement is not an excuse to add new variables.

How a Repair Accidentally Changes the Measured Outcome

Suppose the question asks which setup produces a greater temperature change. Measuring only the time until “it feels warm” changes the outcome from a numerical temperature change to a subjective sensation.

The new procedure may be easier, but it answers a different question.

How a Repair Accidentally Changes the Comparison Population

If the original investigation concerns one type of seed, adding several unrelated seed types may change the scope of the claim.

More specimens strengthen evidence only when the specimens belong to the group the question is actually about.

How a Repair Accidentally Changes the Time Question

Suppose the original question asks about the final amount after 30 minutes.

Measuring every minute is not wrong, but it may introduce a different question about the path or rate of change. If those extra readings require opening or disturbing the system, they can even weaken the original endpoint comparison.

Collect extra evidence only when it serves the claim.

More Apparatus Does Not Mean Better Design

A complex digital sensor can be less useful than a simple ruler if the sensor measures the wrong property.

A stopwatch is irrelevant if the question asks only for final length after the same fixed time.

Science rewards fit between question and evidence, not equipment count.

The Method-Repair Audit

Original question: Write it exactly.
Changed factor: What is deliberately different?
Measured outcome: What response answers the question?
Current weakness: What prevents strong evidence?
Proposed repair: What one change addresses that weakness?
Question-preservation check: Is the changed factor still the same?
Outcome-preservation check: Is the intended outcome still the same?
Alternative-cause check: Did the repair introduce a new factor that could affect the outcome?
Evidence check: What claim becomes more defensible after the repair?

The Earliest-Weak-Link Diagnostic

Failure signature	Earliest weak link	Repair
“Use a different species to get clearer results.”	Improvement adds a new changed factor.	Use suitable similar specimens unless species itself is the intended factor.
“Measure colour instead of growth because colour is easier.”	Measured outcome changed.	Choose a valid measure of the original outcome.
“Add more equipment.”	No method weakness identified.	Name the evidence problem first.
“Repeat it, but use a different setup each time.”	Repeat is not comparable.	Preserve the planned conditions.
“Test much higher values and also change the material.”	Range repair becomes multi-variable redesign.	Extend only the same changed factor first.
“Measure every minute” when opening the container changes airflow.	Measurement repair introduces interference.	Use a less intrusive method or a schedule that preserves conditions.
“Use more samples” from a different population.	Sampling changes the claim scope.	Sample the population the original question concerns.

Misconception Repair — “Any Extra Data Improve the Experiment”

Extra data improve the investigation only if they reduce uncertainty about the original question. Ten precise readings of the wrong outcome are still the wrong evidence.

Misconception Repair — “Changing the Question Is Always Bad”

Scientists often redesign questions deliberately. That can be excellent science. The problem in PSLE method-improvement reasoning is changing the question without noticing while claiming to have improved the original method.

If you want a new question, name it as a new question.

Misconception Repair — “Keeping Everything the Same”

Not literally everything stays the same. The changed factor must vary. What remains comparable are the other relevant factors that could alter the measured outcome.

Misconception Repair — “Better Measurement Can Rescue a Confounded Test”

More precise measurement cannot tell you which of two simultaneously changed causes produced the difference. Repair the comparison architecture first.

How This Appears in Multiple-Choice Questions

Restate the investigation question.
Identify the weakness in the current method.
Check each option for whether it repairs that weakness.
Reject options that introduce a new changed factor.
Reject options that replace the intended measured outcome.
Reject impressive-looking additions that do not address the weakness.
Choose the change that strengthens evidence for the same relationship.

How This Appears in Structured Inquiry Answers

A useful reasoning shape is:

The current method is weak because ______. Improve it by ______ while keeping ______ comparable. This preserves the investigation of how ______ affects ______ and gives stronger evidence about ______.

This is a thinking scaffold, not a compulsory PSLE phrase.

Practice Sequence

Take five investigations and write the question signature.
For each, identify one genuine method weakness.
Propose one small repair.
Invent one tempting “improvement” that actually changes the question.
Explain exactly which part of the question signature changed.
Compare control, measurement, repetition, sampling and range repairs.
Use unfamiliar Science contexts so the decision is not tied to one topic.
Return after several days and repeat without the audit checklist.

Unfamiliar Transfer Challenge

A mystery investigation asks: “How does Condition X affect Outcome Y?” The current method uses one specimen, a coarse measuring scale and two test values.

Which changes preserve the question?

Use a finer suitable scale to measure Y — preserves the question.
Repeat each X condition — preserves the question.
Add intermediate X values — preserves the question.
Use suitable similar specimens if the claim concerns that type of specimen — may preserve the question.
Change X and specimen type together — changes the question.
Replace Y with an easier unrelated outcome Z — changes the question.

The surface Science is unknown. The method-preservation reasoning still works.

Delayed Independent Return

Three to five days later, take a fresh investigation and answer without notes:

What is the original scientific question?
What factor is deliberately changed?
What outcome is supposed to answer the question?
What is the actual method weakness?
What is the smallest useful repair?
Does the repair preserve the changed factor?
Does it preserve the intended outcome?
Does it introduce another possible cause?
What evidence becomes stronger after the repair?

The Answer-Checking Receipt

Did I write the original question before improving it?
Did I identify the changed factor?
Did I identify the intended measured outcome?
Did I name the exact method weakness?
Does my repair address that weakness?
Did I avoid introducing a new changed factor?
Did I avoid replacing the measured outcome?
Did I preserve relevant comparable conditions?
Did I avoid extra apparatus that does not improve evidence?
Can I explain why the improved method answers the same question better?

Evidence and Model Limits

Real scientific research often changes its questions as new evidence appears. Researchers may broaden a population, add variables or introduce new outcomes deliberately. That is not an error.

This guide teaches a narrower Primary Science discipline: when the task is to improve an existing investigation, the learner should first preserve the question being tested and distinguish a repair from a redesign.

There can also be more than one valid improvement. The strongest choice depends on the weakness actually present in the question.

Useful Internal Routes

Parent and Tutor Teaching Guide

When a child suggests an improvement, ask two questions before judging it:

“What weakness are you fixing?”
“Are we still testing the same relationship after your change?”

If the learner cannot name the weakness, the suggestion may be a memorised method phrase. If the learner cannot restate the original relationship after the change, the improvement may have drifted into a new question.

Use near-miss pairs. Present two proposed improvements: one that improves measurement but preserves the question, and one that adds a new changed factor. Ask the learner to explain the difference rather than merely choose.

Then reverse the task. Give a deliberately redesigned investigation and ask the learner to write the new question it actually tests. This helps children understand that redesign itself is not wrong; hidden redesign is the problem.

Return after a delay with a new Science topic. Mastery is shown when the learner protects the question architecture without needing the words “changed variable” or “measured variable” as prompts.

Authoritative and Research References

The research sources support broader inquiry-learning principles. They do not create PSLE marking rules or one compulsory method-improvement sentence.

The Quiet Ending

A scientific question is a promise.

It says: this is the relationship we are trying to understand.

Improve the ruler. Improve the controls. Repeat the test. Add better observations.

But while you repair the method, keep the promise.

Small Group Tutorials